When too many surgery patients come back to a hospital after being sent home, the hospital can be fined by the federal government. But a new study suggests many of those so-called readmissions are not the hospital's fault.

Many readmissions were due to issues like drug abuse or homelessness, the researchers found. Less than one in five patients returned to the hospital due to something doctors could have managed better during the first - or index - hospital stay.

"Very few were due to reasons we could control with better medical care at the index admission," said lead author Dr. Lisa McIntyre, of Harborview Medical Center in Seattle.

McIntyre and her colleagues write in JAMA Surgery that the U.S. government began fining hospitals in 2015 for surgery readmission rates that are higher than expected. Fines were already being imposed since 2012 for readmissions following treatments for various medical conditions.

The researchers studied the medical records of patients who were discharged from their hospital's general surgery department in 2014 or 2015 and readmitted within 30 days.

General surgery includes operations to fix hernias, for example, or to remove a gallbladder or appendix.

Out of the 2,100 discharges during that time, there were 173 unplanned readmissions.

About 17 percent of readmissions were due to injection drug use and about 15 percent were due to issues like homelessness or difficulty getting to follow-up appointments.

More on this...

Only about 18 percent of readmissions - about 2 percent of all discharges - were due to potentially avoidable problems following surgery.

While the results are only from a single hospital, that hospital is also a safety-net facility for the local area - and McIntyre pointed out that all hospitals have some amount of disadvantaged patients.

"To be able to affect this rate, there are going to need to be new interventions that require money and a more global care package of each individual patient that doesn't stop at discharge," said McIntyre, who is also affiliated with the University of Washington.

Being female, having diabetes, having sepsis upon admission, being in the ICU and being discharged to respite care were all tied to an increased risk of readmission, the researchers found.

The results raise the question of whether readmission rates are valuable measures of surgical quality, write Drs. Alexander Schwed and Christian de Virgilio of the University of California, Los Angeles in an editorial.

Some would argue that readmitting patients is a sound medical decision that is tied to lower risks of death, they write.

"Should such an inexact marker of quality be used to financially penalize hospitals?" they ask. "Health services researchers (need to find) a better marker for surgical quality that is reliably calculable and clinically useful."